Image interpretation criteria for FDG PET/CT in multiple myeloma: a new proposal from an Italian expert panel. IMPeTUs (Italian Myeloma criteria for PET USe)

Purpose FDG PET/CT is able to detect active disease in patients with multiple myeloma (MM) and can be helpful for staging and assessing therapy response, but no standard interpretation criteria have been proposed for the evaluation of FDG PET/CT in MM. Methods A group of Italian nuclear medicine physicians and haematologists met to propose new visual interpretation criteria to standardize FDG PET/CT evaluation in MM patients (Italian Myeloma criteria for PET USe; IMPeTUs) and the reproducibility of these criteria was tested. This Italian multicentre protocol was set up as a subprotocol of EMN02, an international prospective multicentre trial of the European Myeloma Network. The criteria were agreed at multidisciplinary consensus meetings. They include a description of the metabolic state of the bone marrow (BM), number and site of focal PET-positive lesions, the number of osteolytic lesions, and the presence and site of extramedullary disease, paramedullary disease and fractures. A visual degree of uptake was defined for the target lesion and extramedullary lesions according to modified Deauville criteria. MM patients who had undergone FDG PET/CT at baseline (PET-0), after induction (PET-AI) and at the end of treatment (PET-EoT) were enrolled. The patients had been prospectively enrolled in EMN02 and their PET scans were a posteriori reinterpreted in a blinded independent central review process managed by WIDENA (R). Five expert nuclear medicine physicians scored the scans according to the new criteria. A case was considered read when four out of the five reviewers completed the report. Concordance among reviewers on different metrics was calculated using Krippendorff's alpha coefficient. Results A total of 17 consecutive patients were enrolled. On PET-0, the alpha coefficients for the BM score, the score for the hottest focal lesion, the number of focal lesions and the number of lytic lesions were 0.33 and 0.47, 0.40 and 0.32, respectively. On PET-AI, the alpha coefficients were 0.09 and 0.43, 0.22 and 0.21, respectively, and on PET-EoT, the alpha coefficients were 0.07, 0.28, 0.25 and 0.21, respectively. BM was generally difficult to score since grades 2 and 3 are difficult to discriminate. However, since neither of the two grades is related to BM myelomatous involvement, the difference was not clinically relevant. Agreement on focal lesion scores and on the number of focal lesions was good.